Information which has been created in the knowledge that it is false in order to influence people; and
Information which is spread in good faith that is correct but contains factual inaccuracy.
While I generally agree with the broader sentiment, I have a few quibbles.
You state under Breaking the Formula and misinformation that “The only way this can happen is through writing on subjects which you have not studied which leaves a knowledge gap.” That is not the only way.
The prime way is that of having a belief that is justified in a set of contexts, and is held as a universal, that is actually simply a context sensitive heuristic.
The very notion of “truth” seems to be a simplistic heuristic in this sense, a shortcut for “very high probability in common contexts”.
The number of classes of systems that are not predictable seems to be infinite.
So we have the very real issue of communicating between individuals with models of reality that are vastly different. Once one understands that what we each experience as reality is most likely actually a subconsciously generated model of reality that is only partly influenced by “objective reality”, and is also influenced by genetic and culturally selected heuristics, experience and choice; then the complexities present expand infinitely. Any sort of communication about a non-trivial subject is a non-trivial problem.
It seems that this reality we find ourselves in is a finely balanced mix of order and chaos. Going too far either way causes problems. Maintaining an appropriate balance seems to be one of the great arts of life (science included).
You state in respect of theoretical physics that it is a “knowledge pyramid”, but to me that conveys the wrong idea. It seems far more probable that it is a potentially infinite journey through levels of systems, algorithms and complexity. There do in fact seem to be an infinite set of classes of computational systems, some of which are predictable and some of which are not.
The idea of a pyramid with a top doesn’t do justice to the underlying data-structure, however accurately it may represent the embodiment of knowledge in living human beings.
I tend to use the rule that if I don’t look up the details then I am acting irresponsibly, because there are just too many details for anyone to be comfortable with. I have a large array of heuristics I use for 1st order approximations when doing scenario analysis, but when it gets down to detailed design, then I ensure I check everything, with multiple independent sources if possible, and if it seems appropriate and important enough, do the work myself.
Peter Berg has done a good job on my final quibble, so I won’t elaborate further.